This project brings to completion data access, organization, documentation, coding, entry, analysis and dissemination of results from an existing prospective longitudinal data set. The data focuses on a cohort of children born in the greater Helsinki area between July l, 1975 and June 30, 1976. Through the use of Finnish national registries, and temperament assessment probes, it allows for the exploration of relations between maternal pregnancy stress, maternal depression, and neonatal delivery status on the temperamental characteristics of the child. Further, it allows for the exploration of relations between maternal ratings of infant temperament (6-8 months) and preschool temperament (5 years), and teacher ratings of temperament in early adolescence (age 12), and behavior problems/mental illness during childhood and adolescence. Behavior problem data are available from mothers, school nurses, and teachers. Mental illness is assessed through physician diagnosis at hospitalization, with records available over the life-span of the child. Important mediating variables also studied include family stress (financial and medical, psychiatric problems of siblings, parents, and grandparents), developmental rate of the child during the preschool years, somatic illness of the child from age 1 through 16, and educational outcomes in preadolescence. Major longitudinal questions will be analyzed using the partial least squares approach to model testing. Due to large sample sizes (800 to 2100 for various measures) and the range of data available from infancy through age 16, this data set allows for the most extensive analysis of long-term temperament/mental illness relations currently available.